ENHANCING HFO SEPARATOR EFFICIENCY: A DATA-DRIVEN APPROACH TO PETROLEUM SYSTEMS OPTIMIZATION

Abstract

The study titled Enhancing HFO Separator Efficiency: A Data-Driven Approach to Petroleum Systems Optimization investigated the quantitative relationships between thermophysical, mechanical, and operational parameters that determine the performance efficiency of heavy fuel oil (HFO) centrifugal separators. Using a data-driven framework, the research analyzed empirical field data supported by an extensive review of 112 scholarly and industrial studies focusing on petroleum separation technologies, machine learning applications in process optimization, and reliability-based maintenance modeling. The quantitative methodology incorporated multiple regression and mixed-effects models to evaluate the influence of inlet temperature, viscosity, flow rate, bowl rotational speed, torque, and vibration amplitude on overall separator efficiency, measured through a composite Efficiency Index (EI) integrating residual water, solids concentration, and energy consumption per ton of fuel treated. The findings revealed that inlet temperature had a significant positive impact on separator efficiency, whereas viscosity, torque, and vibration amplitude exhibited strong negative effects. Bowl rotational speed contributed positively but displayed diminishing marginal returns beyond optimal operational thresholds, while flow rate had minimal direct influence within controlled capacity ranges. The study demonstrated that approximately 75% of the total variation in separator efficiency could be explained by the identified thermophysical and mechanical predictors, confirming that the process is jointly governed by thermal conditioning and mechanical stability. Reliability and validity testing confirmed high measurement accuracy, while multicollinearity diagnostics verified the statistical independence of predictors. The analysis also established that mechanical imbalance, characterized by elevated torque and vibration, directly contributed to reduced efficiency, supporting the integration of condition monitoring and predictive maintenance as essential optimization strategies. The review of prior literature reinforced the study’s empirical findings, revealing convergence between historical theoretical models and modern data analytics in understanding separator behavior under variable operational conditions. Overall, the study concluded that HFO separator efficiency can be substantially enhanced through a unified data-driven optimization framework that integrates real-time monitoring, predictive analytics, and reliability engineering. This approach provides a comprehensive quantitative foundation for petroleum system operators to improve energy efficiency, minimize contaminant carryover, and extend equipment lifespan through adaptive, analytics-based decision-making.